Automatic battles: Final conclusion

[drus45]

Ok I tell you guys all I learned in my testing of auto-battles (and I test with medium difficulty).

At least these factors affect auto-battle results:
-hit_points = medium effect
-hit_point_extra = medium effect
-troop numbers = biggest effect
-primary melee attack = big effect
-ranged attack (if unit has ranged attack then also melee attack has effect) = big effect
-armor = big effect

I don't know the mathematical operation the computer makes to calculate the auto-battle results. It seems that some factors are more important than others, so thats why I put these "medium", "big", "biggest". I might even be wrong about which one of those matters most, but I think I'm right.
Also, for example, hypaspist (with 160 men) wins gold shields and loses to elite hoplite but the gold shields win the elite hoplite. So there is not one way all these things are calculated. It means things like phalanx, heavy, light or missile type troops might have some effect when facing certain opponents.

But the main thing is that currently in Extended Realism Mod the automatic battles are flawed.

Some of the most powerful units at the moment are:
-aor numidia spearmen (240 men, powerful ranged attack and medium melee attack)
-numidian javelinmen (240 men, powerful ranged attack and medium melee attack)
-aor asia roman archer (medium ranged attack, medium melee attack and +8 armor)
-german warband (240 men, powerful ranged attack and medium melee attack)

Some of the weakest units are:
-hypaspists (very few men)
-triarii (very few men)
-all cavalry units (very few men and hit_point_extra don't work with them)

Other weak units:
-legionares
-sacred band
-silver shields
-elephant units

This means that every single unit must be tested in auto-battles and edited. And some units in battle map are also underpowered (hypaspists, triarii, silver shields), mainly because of low number of men.

Other mistakes I have noticed:
-There is problem with hoplites whose formation is: formation 0.7... They are too close together and even when highly_trained their formation is disorderly.
-The german berserkers have javelin instead of sword in their hand in battle map mode. This is because their javelin weapon have been deleted from export_descr_unit, but their skeleton still has it. I don't know how you can take the javelin away without them using it in battle map mode. So I suggest to bring back their javelin.
-Most of the horse sprites do not match with their (new ExRM) textures.
-Some units (carthagian citizen peltast, hypaspist swordsmen) don't have any text in their info box.

Other things that might or might not be mistakes:
-Macedonian Pezhetairoi have darker skin than what I think ancient macedonians had
-Spartan hoplites also have darker skin than what I think ancient greeks had.
-Britons have too bright (blue) colored clothes.
-merc elephants cannot be hired from anywhere.

I will have to say this mod in its current state is UNPLAYABLE. The auto-battles must be balanced and I say all other improvements right now are not important. Balancing the auto-battles will have big effect on AI behaviour and the succes of all factions. There is no point right now for example test what the AI will do in the campaign map(will ptolemy attack greece or seleucid etc...) because all this will change and all testing must be done again.

Lots of work ahead.

[Arkaeyn]

I will have to say this mod in its current state is UNPLAYABLE.

I'm playing and enjoying the mod right now, and have been for a few weeks. Not perfect doesn't mean not playable.

The rest of your post was entirely reasonable. Being insulting and excessively dramatic at the end is likely to ruin that.

[despot_of_rhodes]

I haven´t posted here for a while, since my RTW CD broke.. but i was trying to keep up with news.

From what i noticed, i don´t think drus45 was about to insult anyone, it´s maybe just his engagement and his passion, trying to make his point of view, which may appear as such. I think that his tests concerning autoresolve situation were at least intensive. And i also think, there is no need, to use word like "insult" to someone, who´s is trying to help, maybe with a wrong choice of words

But back to the topic: If autoresolve battles ( not only between human and AI, but also among AI factions ) are unrealistic, so to say, then the whole historicaly accurate concept of trying to recreate the ancient world in that timeframe might be corrupted as well. That is the main purpose of this thread, i think. So, maybe Quinn should take a look at this and say smth.

Greetings,
despot.

[Arkaeyn]

It's not the first time he's done this. He comes storming into this particular mod's forum, making extreme declarations about how this mod, and this mod alone, does not meet his auto-resolve needs. This ignores the facts that:

1. Every version and mod of RTW has ahistoricism issues, particularly in regards to which factions become superpowers. So why focus on this mod in particular?
2. Every version and mod of RTW has significant differences between auto-resolve results and played battle results. Again, why focus his ire on this mod?
3. Not every RTW player considers auto-resolve results to be the most important aspect of the game.
4. Presentation matters. I would love it if the auto-resolve results could be brought into equilibrium with regular battle results. Drus42, apparently, would love it as well. So why do I roll my eyes at the end of that first post? I'm a natural ally.

[DraeKlae]

Having triarii with less men is historically accurate, because the romans used triarii maniples with half the manpower of hastati or principes maniples. Regarding the Hypaspistai, they were an elite guard intended to protect the macedon king, although Alexander used them extensively in decisive manuevers (somewhat similar to the extraordinarii for the romans).

Besides that, I don't believe the field battles should be unbalanced in favor of the auto-resolve battles by changing stats and adding manpower to certain units. For example, you just can't put 200 men (huge unit size) in a sacred band unit; they'd become too powerful.

[drus45]

It's not the first time he's done this. He comes storming into this particular mod's forum, making extreme declarations about how this mod, and this mod alone, does not meet his auto-resolve needs. This ignores the facts that:

1. Every version and mod of RTW has ahistoricism issues, particularly in regards to which factions become superpowers. So why focus on this mod in particular?
2. Every version and mod of RTW has significant differences between auto-resolve results and played battle results. Again, why focus his ire on this mod?
3. Not every RTW player considers auto-resolve results to be the most important aspect of the game.
4. Presentation matters. I would love it if the auto-resolve results could be brought into equilibrium with regular battle results. Drus42, apparently, would love it as well. So why do I roll my eyes at the end of that first post? I'm a natural ally.

The reason I'm "storming" here is simple: this is the best mod available. And I don't want to play anything else. I have also said that this problem is with every mod, maybe even with vanilla RTW. But I'm not going to test with vanilla, because I personally think that RTW without mods is a horrible game. Most of all the graphics and colors are awful in vanilla game.

I actually meant to say (and I think I said) that this auto-resolve thing is more important than any faction becoming superpower, because changing the auto-resolve will change the succes of every faction. And like despot said, even if the player is not interested in using the auto-resolve, the AI uses it every single time it fights with itself. And more importantly, there are people (most importantly me) who really don't want to fight every battle on the battle map.

And about the mod being unplayable... well it is a war game and there are two ways of deciding battles and the other way is flawed, then the game is flawed. I just want to make a point, which is very important. My previous messages didn't get the attention this matter needs. Maybe it's good because there were some errors with my calculations, and I wrote too early when I was still testing.

And about the presentation... I have searched the whole Total War Center and there is very little talk and no productive talk at all about auto-resolve. This I find strange because I have always found the auto-resolve to be the weakest link in the whole series. Comparing it for example to Hearts of Iron 2 battles makes RTW auto-resolve look totally retarded. HOI2 takes into account: ground, weather, time (night), COMBINED ARMS and all the regular stuff like experience, leader, attack and defence etc. All RTW auto-resolve is interested is hit points (which are not really used at all), hit point extra (which is only used with elephants and chariots, and I think is not even supposed to have any effect on infantry), leader, troop numbers (wow, really?), attack (how clever) and armor (how do they come up with all these great ideas?).

And by the way I think the new game, Empire TW is the worst in the series and I haven't even played it. It's like they are getting crappier all the time. Troop numbers are so low that it's like playing some stupid real time "strategy" puke. But the graphics and details are soo great. Well let me tell you something; they look like and I don't even care about the graphics that much. I want more numbers in battles, otherwise the whole series look like some stupid low budget film or some gay fantasy role playing game with elves on it. And the naval battles are the worst idea to spend all your time developing the game. People in general are totally out of ideas.

OK maybe I'll take a break, I think I just proved at least on point Arkaeyn made. But my intention is to make this mod better.

[Arkaeyn]

All right. Deep breaths, drus! You're on the right track, just take those deep breaths. There's a piece of advice that floats around that says that if you want to make constructive criticism (ie, convince someone to change when they don't have to), you put the criticism between two compliments. Saying "this is the only mod I want to play" is right along those lines.

And again, there's a difference between unplayable and flawed. Every game is flawed, except perhaps radically simple games like Bejeweled or Tetris, and I'd still say they're flawed. If this is the best mod available for a game you play, then unless the flaw is game-breaking, it's still playable. Actually, I tend to believe that most pieces of media have flaws, indeed, require some flaws, in order to demonstrate the quality of the rest of the media. I remember a classic game like, for example, Deus Ex for its flaws because they went hand-in-hand with its strengths.

You may be right about the auto-resolve not being discussed enough. But it may also be that the auto-resolve is hard-coded and can't be fixed, and/or changing unit stats in order to achieve "soft" fixes breaks the battle engine. And again, I'd far prefer a working battle engine to a working auto-resolve engine. When I think of memorable RTW experiences, I think of the desperate battles I had against hordes of Ptolemaic hoplites and elephants as Pontus, facing the pride of an empire and throwing them back repeatedly. I don't think of the Grand Campaign map because it's simply not as interesting as the battles. It's a means, for contextualizing the battles, far more than an end. A game like Civilization IV does grand strategy far more effectively.

And saying that you think that Empire is the worst of the series without having played it...? I haven't played it. But then, I think that, with the exception of Rome, the Total War series in general is rated too highly (I also haven't played Medieval 2, so I could be wrong). Shogun was far too dense, I had no idea what was going on and what tactics or strategies were effective. Medieval was quite the opposite, far too easy. So it's not like you're attacking sacred cows for me here.

People in general are totally out of ideas.

This is the kind of thing I said 10 years ago. Innovation is nice, but only goes so far. And people have always been out of ideas. But ideas are not the problem here. The idea of the Total War games is not fresh or new. Strategy games have been combining simpler campaigns with complex tactics for decades. The problem is the implementation - in your case, the auto-resolve is ineffective and ahistorical. But this is a focus on numbers. Brilliant ideas won't make it better, hard work might, but it also might not. And we ain't getting paid for this.

[Bleakmage]

I don't want to play the new one either. This is also the only mod I play, until Total Realism 7 comes out. I like these ancient battle games. Sea battles in Roman times would have been fun. Sea battles in the new game has already been done many times. I want to play something like Hyborian Age: Total War. It would have alot in common with vanilla TW, I think.

[Rameth]

I'm not really sure what you guys are trying to accomplish by arguing but shouldn't you try to stay on topic?

Now before I even go onto what I was going to ask I first would like know is it even possible to change the stats for autoresolve without unbalancing the field battles? I mean if you want to keep unit numbers historical but unit size is the biggest impact on autoresolve then theres a serious problem with trying to fix it before even going into anything.

How would you actually get past this problem?

[Newtothegame]

I don't know anything about modding. I don't know if it is hardcoded. If it wasn't, you could let unit properties like HP_extra or whatever it is called determine the outcome of autobattles. In field battles that particular property shall be unimportant. I don't know, whether or not the whole autocalculation of battles can be made using one quality to a huge and the others to a minimum extent.

If you dedicated some unit properties to only effect autobattles, it won't spoil the battle map fighting.

[Scutarii]

Well, priority one: balance field battles. Hoping, of course, that it spill over to a balance in autobattles.

When that fails, balance secondary HP to tweak results a little. That's the most Quinn can really do without focussing entirely on auto-battles. Perhaps try making some of the changes you proposed yourself (and then releasing a sub-mod for it). The problem is that most people tend to fight their battles, not auto-resolve them, as its one of the TW series biggest features, so balancing things just for auto-resolve (like raising numbers of Hypaspists or something because they're inneffective in auto-resolve) wouldn't make sense in field battles, so it doesn't happen.

I think the reason that these missile troops with decent melee stats are so effectivein auto-resolve is because many troops in ExRM have very low armour values. So missiles do a lot of damage.

[Ichon]

I actually think this is interesting thread because the auto resolve is what the AI uses when fighting armies between its selves. In another thread talking about how to get Parthia to not die so early squeezed between Bactria and Sels without making it overly aggressive could bear on this as well as a couple other factions.

Its not about solving it just so a human player can get out of fighting a battle since most people aren't going to do that very often until late in the game when even if autoresolve poorly reflects if player controlled battle, the resource are there to use it just to take the final 10 regions to win a campaign or completely occupy world map.

[Vae victis]

I think this mod is very playable. The AI expands well and challenges the player. Battles are fun to play out on the field. Thats what makes total war games great. If you want to auto-resolve all of your battles, but still want some strategy, try some Paradox titles. I have been playing this mod for a while and it is very playable, and very enjoyable